基于最优模型的0.1°全球植被带逐月潜在蒸散发数据集

在气候学、生态学、水文学和农学等多个领域，选择有效模型来准确估算潜在蒸散发（PET）至关重要。然而，当前许多主流的PET产品依赖于使用默认参数的模型，这会给PET相关研究带来不确定性。为解决该问题，我们基于全球124个通量塔的观测数据，推导了五种常用PET模型的参数。通过比较这些模型在不同生物群系中的性能，我们确定校准后的Priestley-Taylor模型和Milly-Dunne模型为最优选择，它们具有很好的空间可移植性，能够有效应用于其原始观测站点之外。利用这些最优模型，结合四个广泛使用的气象数据集以及随时间变化的土地利用/覆盖数据，我们生成了1992-2022年的0.1°全球植被带逐月潜在蒸散发数据集。这个新的PET数据集为开展PET相关研究提供了一个新的选择。Accurately estimating Potential Evapotranspiration (PET) with effective models is critically important across multiple disciplines including climatology, ecology, hydrology and agronomy. However, most current mainstream PET products rely on models with default parameters, which introduces uncertainties into PET-related research. To address this issue, we derived the parameters of five widely used PET models based on observations from 124 global flux towers. By comparing the performance of these models across different biomes, we identified the calibrated Priestley-Taylor and Milly-Dunne models as the optimal choices, which exhibit excellent spatial transferability and can be effectively applied beyond their original observation sites. Using these optimal models combined with four widely used meteorological datasets and time-varying land use/cover data, we generated a global monthly Potential Evapotranspiration dataset at 0.1° spatial resolution for various vegetation zones spanning the period from 1992 to 2022. This newly developed PET dataset provides a novel alternative for PET-related research.

搜集汇总

背景与挑战

背景概述

该数据集是一个全球植被带逐月潜在蒸散发（PET）数据集，旨在通过基于124个通量塔观测数据校准的最优模型（Priestley-Taylor和Milly-Dunne模型）来减少PET估算的不确定性，覆盖1992年至2022年，空间分辨率为0.1°，以月为时间单位提供开放获取的NetCDF格式数据。